There are several unpaid duties that go with being a scientist–among them, refereeing manuscripts for publication, serving on grant review “study sections” and of course writing recommendations. This last obligation is often the least understood.
There are several aspects to writing a recommendation letter that are clearly fraught with ambiguity:
First, given our lawsuit-friendly times, how frank should one really be? My opinion is that if you can’t write someone an outstanding support letter, it’s really better to tell them that up front so that they can go elsewhere. I realize that this might put other institutions and granting agencies at risk, but, following my principle, a really awful scientist is going to have a hard time getting any support letters.
Second: the phenomenon of “recommendation letter inflation”. It’s just a fact of life–the intense competition for jobs and grants probably pushes the trend, but if you really want the individual to get the job, you’d better not “condemn them with faint praise”. This goes against the grain for a lot of scientists, because in their papers, when discussing their data, any tendency towards over-hype is very properly held against one.
Third, it’s key to get across in the very first or second sentence, how long you have known the person you are recommending –and in what capacity. If this isn’t made clear, the entire letter becomes considerably less useful to the reader.
Finally, I always believe that it’s useful to step back at some point and indicate where the individual’s trajectory appears to be headed in your opinion– are they headed for the National Academy? Or maybe just towards being a solid scientific citizen (well funded and publishing in good journals).
Jim
Discoverers and Discoveries 